Tanzu Kubernetes Cluster Example Deployments

Jun 23, 2023 · 2 min read · VMware Tanzu Tanzu Kubernetes Clusters yaml ·

Intro

This post gives a few examples of Tanzu Kubernetes Cluster Manifests and how to deploy them.

Prerequiries

Successfull Installation of vSphere with Tanzu Supervisor
vSphere Namespace created
Already existing VM Classes, Storages Policies and Tanzu Kubernetes Releases assigned to the Namespace.

You will find the IP to your Kubernetes API on the Namespace Option "Link to CLI Tools"

1kubectl vsphere login --server=10.40.80.20

1kubectl config get-contexts

1kubectl config use-context <>

Get Parameters of your vSphere Namespace

VM Class
Storagge Policy
Tanzu Releases

1kubectl get vmclass

1kubectl get storageclass

1kubectl get tanzukubernetesrelease

Cluster with 3 Control and 6 Worker Nodes

 1apiVersion: run.tanzu.vmware.com/v1alpha3
 2kind: TanzuKubernetesCluster
 3metadata:
 4 name: tkgs-cluster
 5 namespace: tkgs-cluster-ns   
 6spec:
 7  topology:
 8   controlPlane:
 9    replicas: 3
10    vmClass: best-effort-large  
11    storageClass: workload-management-storage-policy
12    tkr:
13      reference:
14       name: v1.23.8---vmware.2-tkg.2-zshippable
15   nodePools:
16   - replicas: 6
17     name: worker
18     vmClass: best-effort-large  
19     storageClass: workload-management-storage-policy

Save this file as a yaml file on your workstation.

Apply Cluster Manifest

1kubectl apply -f clusterspecs.yaml

1kubectl get tanzukubernetescluster

Check the Status of the Cluster. "READY" should be true after some minutes.

Edge Cluster

 1apiVersion: run.tanzu.vmware.com/v1alpha3
 2kind: TanzuKubernetesCluster
 3metadata:
 4 name: tkgs-cluster
 5 namespace: tkgs-cluster-ns   
 6spec:
 7  topology:
 8   controlPlane:
 9    replicas: 3
10    vmClass: best-effort-small  
11    storageClass: workload-management-storage-policy
12    tkr:
13      reference:
14       name: v1.23.8---vmware.2-tkg.2-zshippable
15   nodePools:
16   - replicas: 3
17     name: worker
18     vmClass: best-effort-medium  
19     storageClass: workload-management-storage-policy

Minimal Cluser for PoC

Use this cluster only for PoC/testing. Only three VMs are created in total. Later, you'll use Kubernetes Taints to enable the Control node to run user workloads. This is useful when testing workloads with a minimal VM footprint when deploying in environments with few resources available.

Concept of Kubernetes Taints In Kubernetes, taints are a way to mark a node with a special attribute that affects which pods can be scheduled onto that node. Taints are used to repel or attract pods based on certain criteria, such as node characteristics or hardware specifications.

 1apiVersion: run.tanzu.vmware.com/v1alpha3
 2kind: TanzuKubernetesCluster
 3metadata:
 4 name: tkgs-cluster
 5 namespace: tkgs-cluster-ns   
 6spec:
 7  topology:
 8   controlPlane:
 9    replicas: 3
10    vmClass: best-effort-medium  
11    storageClass: workload-management-storage-policy
12    tkr:
13      reference:
14       name: v1.23.8---vmware.2-tkg.2-zshippable
15

Do not use this in a production environment After a successfull deployment of the cluster run the following commands to allow the Control Nodes to run User Workloads:

1kubectl taint nodes --all node-role.kubernetes.io/master-